Expression of macrophage inflammatory protein-1α (MIP-1α) in human endometrium throughout the menstrual cycle


  • Minoru Akiyama,

    Assistant Professor , Corresponding author
    1. Department of Obstetrics and Gynaecology, Shiga Universiry of Medical Science, Japan
      Correspondence: Dr M. Akiyama, Department of Obstetrics and Gynaecology, Shiga University of Medical Science, Setatsuki- nowa-cho, Otsu, Shiga, 520–2192, Japan.
    Search for more papers by this author
  • Hidetoshi Okabe,

    1. Department of Clinical Laboratory Medicine, Shiga Universiry of Medical Science, Japan
    Search for more papers by this author
  • Kenji Takakura,

    Associated Professor
    1. Department of Obstetrics and Gynaecology, Shiga Universiry of Medical Science, Japan
    Search for more papers by this author
  • Yoshihide Fujiyama,

    Associated Professor
    1. Second Department of Internal Medicine, Shiga Universiry of Medical Science, Japan
    Search for more papers by this author
  • Yoichi Noda

    1. Department of Obstetrics and Gynaecology, Shiga Universiry of Medical Science, Japan
    Search for more papers by this author

Correspondence: Dr M. Akiyama, Department of Obstetrics and Gynaecology, Shiga University of Medical Science, Setatsuki- nowa-cho, Otsu, Shiga, 520–2192, Japan.


Objective To investigate the distribution and the role of macrophage inflammatory protein-la (MIP- 1α) in human endometrium during cyclic changes.

Setting Department of Obstetrics and Gynaecology, Shiga University of Medical Science and University Hospital.

Materials Eighteen endometrial tissue specimens surgically resected or biopsied from women with normal menstrual cycles, without hormonal disorder or endometrial diseases.

Methods By immunohistochemistry, using monoclonalantibodies (lambda delta 78 for MIP-la and CR3/43 for human leukocyte antigen-DR (HLA-DR)).

Results Immunoreactivity for anti-MIP-lα was distributed diffusely in epithelial cells throughout the proliferative and secretory phases but was absent during menstruation due to degenerative or necrotic changes. HLA-DR was expressed in epithelial cells only in the late secretory phase and was not expressed in stromal cells.

Conclusion Immunohistochemicalanalysis showed the presence of MIP- lα in the endometrial epithelium. Expression of HLA-DR in epithelial cells was observed only in the late secretory phase, suggesting that accumulation of MIP-lα in epithelium occurred by self production and not via a receptor mediated pathway. MIP-lα was released from the denuded epithelium during menstruation and appeared to contribute to the accumulation of monocytes/macrophages into the endometrial cavity. MIP-la has a number of biological effects other than monocytic chemotaxis, and some of these effects may be exerted in the endometrial tissue.


Chemokines are a group of inflammatory cytokines with similar amino acid sequences, that are important in the regulation of inflammatory cell recruitment. Chemokines are mainly classified as members of the C-C or C-X-C family, depending on whether the first two cysteines in the motif are adjacent (C-C) or are separated by an intervening amino acid residue (C-X-C). C-X-C chemokines induce chemotaxis of neutrophils, and members of the C-C family generally have chemoattractive effects on monocytes. Both C-X-C chemokines, such as interleukin-8 (L-8)1, and C-C chemokines, such as monocyte chemotactic protein- 1 (MCP- 1)2, are synthesised by endometrial tissue. The synthesis of these chemokines as wellas several other cytokines in the endometrium is upregulated during the secretory phase. These agents are believed to be involved in the recruitment of inflammatory cells, which scavenge endometrial tissues during menstruation3,5. However, the expression of macrophage inflammatory protein-la (MIP-la), a C- C chemokine, endometrial tissue has not been investigated. MIP-1α was originally described as an endotoxin-induced macrophage product with neutrophile chemoattractant and activating properties6. However subsequent studies have shown that the target cells of this chemokine are T cells and macrophages7. We used immunohistochemistry to examine the expression of MIP-1α throughout the menstrual cycle in endometrial samples obtained from nonpregnant women. We also examined expression of human leukocytes antigen (HLA)-DR (CR3/43) to clarify the mechanism of MIP-1α expression since it may be required as an auxiliary factor for the receptor mediated uptake of this chemokine, and discuss about its role on the regulation of inflammatory cell recruitment during the menstruation.


We examined expression of MIP-la in surgically resected (n= 10) and in biopsied (n= 8) endometrial tissue specimens obtained at Shiga University of Medical Science Hospital (1995 to 1996) from women with normal menstrual cycles. The women had required hysterectomy for diseases other than endometrial dysfunction such as early cervical cancer. These women had no hormonal disorders and no ovarian tumours, and none had a history of hormone replacement therapy. Clinical cyclic dating of the endometrium was performed according to the standard criteria8. Three biopsy specimens were taken from the endometrium during menstruation and three specimens (one biopsy and two resected specimens) were obtained in each of five remaining phases: the early and late proliferative phases and the early, mid and late secretory phases. Specimens were embedded in paraffin after being fixed in 10% for- malin, cut into 4–0 pm sections, and processed for light microscopy and immunohistochemistry. The primary antibodies used are listed in Table 1. The anti-MIP-la monoclonalantibody (lambda delta 78), which recognises an antigen epitope that is stable after formalin fix- ation, was developed at our institution9.

Table 1.  Source and dilution of primary antibodies used. MIP-la = macrophage inflammatory protein-la; HLA-DR = human leukocyte antigen-DR.
CD3Anti-pan T cell1:100IgG, polyclonalDako Co Ltd
CD8Anti-killer/suppresser T cell1:50IgG, polyclonalDako Co Ltd
LysozymeAnti-lysozyme1:200IgG, polyclonalDako Co Ltd
HLA-DRCR3l431:100IgG, polyclonalDako Co Ltd
MIP-1αLambda delta 781:2500IgG, polyclonalDako Co Ltd

Sections were treated for 10 minutes with 0.3% hydrogen peroxide in methanol to block endogenous peroxidase activity and then incubated with the primary monoclonalantibody at 4°C for 10 to 12 hours. Immunostaining was performed by the step streptoavidine biotin peroxidase method (Dako Japan Co, Kyoto, Japan) according to the manufacturer's recommendations. Peroxidase activity was detected using the chromogen diaminobenzidine (Dako Co). Nuclei were counterstained with methyl green. Sections for immunostaining of MIP- 1 a, lysozymes, and CD3 posi- tive T cells were pretreated with 0.1% trypsin in 0.2% calcium chloride for 15 minutes at 37°C after deparaffinisation. Sections for CD8 staining were autoclaved in 0.01 mmom of phosphate buffer for 1 minute at 125°C. Macrophages within endometrial tissue were used as the positive internal control for MIP-la. Negative control slides were prepared by incubating sections with phosphate buffered saline instead of the primary antibody. To investigate the possibility that expression of this chemokine was receptor mediated, we also studied the expression of human leukocyte antigen (HLA)-DR (CR3/43), which may be an essential cofactor of receptor mediated MIP-la expression10.


Histological cycle dates of our 18 endometrial specimens were consistent with the dates established by changes in body temperature. Immunoreactivity for MIP- la was diffusely distributed in the cell bodies of glandular epithelium in all specimens taken during the cycle phases (Table 2 and Fig. 1). Afew MIP-la positive cyto- plasmic granules were observed in endometrial stromal cells when they underwent decidual changes but the staining was very weak in the remaining phases. A few inflammatory cells were scattered in the stroma during the proliferative to early secretory phases. The majority of these inflammatory cells were composed of lysozyme-positive mono-histiocytes, and CD3 and CD8 positive T lymphocytes. Expression of these two kinds of inflammatory cells increased gradually during the mid to late secretory phases. MIP- la was strongly expressed in macrophages and in some T lymphocytes. Immunoreactivity for MIP- 1 a was markedly reduced or absent from the degenerated endometrial glandular epithelium in the premenstrualand menstrual endometrium (Fig. 2); lysozyme-positive scavenger macrophages were abundant in the stroma; HLA class II antigen (HLA-DIUCR3) was strongly expressed in macrophages throughout the menstrual cycle, but was not expressed in endometrial epithelium or stromal cells during the early proliferative to mid secretory phases. However, immunoreactivity for HLA-DR gradually increased in endometrial glandular cells during the mid to late secretory phases and was strong in the stroma with an increase in the populations of phagocytes and T lymphocytes (Fig. 3). Endometrial stromal cells did not exhibit immunoreactivity for HLA- DR in any phase of the menstrual cycle.

Table 2.  Immunohistochemistry of macrophage inflammatory protein-1α (MIP-1α) in human endometrium. ++= > 75% positive staining; += > 25% and < 75% positive staining; ?= < 25% and/or few stained cells count; –= no staining.
 Proliferative phaseSecretory phase
  1. *Lysozymes, and CD3 and CD8 positive inflammatory cells infiltrated > 10 per high power field by light microscopy.

Glandular epithelium in functional layer++++++++++
Glandular epithelium in basal layer++++++++++
Stromal cells in functional layer+*±±+*+*
Stromal cells in basal layer+*±±±+*
Figure 1.

Immunostaining for macrophage inflammatory protein-1α (MIP-1α) in the functional layer: (a) In the late proliferative phase, glandular epithelium showed strong positive stain with nuclear mitoses. Stromal cells had shown very weak staining only a few positive cells (original magnification × 400); (b) In the early secretory phase, positive staining was greatest on the surfaces of the lumens in the glandular epithelium and cells were characterised by subnuclear vacuolae. However staining was very weak or absent in the stroma (original magnification × 400); (c) In the mid secretory phase, the glandular epithelium showed positive staining and the staining pattern was similar to that in the proliferative phase (original magnification × 400); (d) In the late secretory phase, both the glandular epithelium and stromal cells showed positive staining and many infiltrating inflammatory cells also showed strong positive staining (original magnification × 400).

Figure 2.

Immunostaining for macrophage inflammatory protein- 1α (MEP-1α) in the menstrual phase: The glandular epithelium with degenerative change showed negative staining and stromal cells and infiltraing imflammatory cells showed very strong positive staining (original magnification × 400).

Figure 3.

Immunostaining for CD8 in the functional layer: Numerous CD8 positive phagocytes and T lymphocytes infiltrated the endometrial stroma and invaded the glandular epithelium (original magnification × 200).


Menstruation has a striking similarity to the processes of inflammation and wound repair, except that the formation of granulation tissue or scar tissue does not contribute to the remodeling and repair of the denuded endometrium. Neutrophils and phagocytes accumulate to scavenge the collapsing endometrial tissue. Chemokines, a group of proinflammatory cytokines, are believed to have an important role in this process. IL-8, a member of the C-X-C family produced by endometrial perivascular stromal cells I, induces selective chemotaxis and activation of neutrophils and has been postulated as the signal for neutrophil recruitment. MIP-1α belongs to a different subclass of chemokine, the C-C family, and has a potent chemoattractive effect on macrophages. MIP-la has been demonstrated in several types of human epithelial tissue and activated mesenchymal cells15–16, but the distribution of this chemokine in endometrial tissue has not been previously examined. To our knowledge, this is the first report to demonstrate that MIP-1α is distributed in human endometrial epithelial cells throughout the menstrual cycle and disappeared from epithelial cells when they underwent degenerative change. Therefore this chemokine appears to be released from epithelial cells during menstruation leading to the recruitment of macrophages into the uterine cavity. Accumulated neu- trophils and/or macrophages are then activated by other cytokines produced by endometrium, such as IL-15, IL- 617, granulocyte colony stimulating factor (G-CSF)5, granulocyte macrophage colony stimulating factor (GM-CSF)18, leukemia inhibitory factor (LIF)19–21 and tumour necrosis factor (TNF)-a protein22, and contribute to degradation and elimination of tissue debris.

Inflammatory cells were not abundant in the endometrium, except during the menstrual phase. However, several cytokines derived from endometrial tissue are chemoattractants for T lymphocytes and promote their accumulation23. In the endometrium CDS positive cells were dominant in the T cell fraction24. MIP-la has chemoattractive effects for CD8 positive lymphocytes, in addition to mono-histiocytes25 and thus may contribute to the accumulation of T lymphocytes. MIP-la is incorporated into the target cells by its receptor, and HLA-DR (CR3) may be required as an auxiliary factor for this receptor mediated process10. However, expression of HLA-DR in the endometrial epithelium is restricted to the mid-secretory to premenstrual phases according to previous reports 26–27, and its expression was not relevant to MIP- 1 a immunoreactivity in the present study. Accordingly, MIP-1 a, like several other endome- trial cytokines appeared to have been synthesised by endometrial epithelial cells. Synthesis of cytokines by the endometrial epithelium is upregulated during the late secretory phase. Secretory changes induced by progesterone administration have been found to upregulate the synthesis of several cytokines and to promote subsequent accumulation of leukocytes in the stroma. MIP-1α synthesis also seems to be involved in this process.

Some cytokines secreted from the endometrial epithelium appear to have autocrine effects on endometrial tissue. For example, an increase in TNF-1α apparently causes a decrease in bcl-2 gene products and induces apoptosis in endometrial epithelial cells during the menstrual shedding28–30, although Fas also seems to be involved in this process31. MIP-1α has several biological effects other than monocytic chemotaxis and some of these activities may be exerted in endometrial tissue via the receptor mediated pathway. HLA-DR (CR3/43) may be required for receptor mediated uptake of MIP-1α, but endometrial stromal cells did not show significant expression of this antigen in the present study. Endometrial epithelial cells strongly expressed HLA-DR during the late secretory to premenstrual phases, probably due to the effects of cytokines derived from activated T cells in the local environment32, suggesting that MIP-1α may have some autocrine effects in epithelial components.

The present results suggest that MIP-1α is synthesised in endometrial epithelial cells and is involved in several important phenomena during the menstrual cycle. Further studies are needed to examine the distribution of its receptor to clarify the exact function of MIP-1α in the endometrium.